The purpose of this report is to evaluate different methods for identifying states in robot models. Both linear and non-linear filters exist among these methods and are compared to each other. Advantages, disadvantages and problems that can occur during tuning and running are presented. Additional measurements from accelerometers are added and their use with above mentioned methods for state estimation is evaluated. The evaluation of methods in this report is mainly based on simulations in Matlab, even though some experiments have been performed on laboratory equipment.The conclusion indicates that simple non-linear models with few states can be more accurately estimated with a Kalman filter than with an extended Kalman filter, as long as only linear measurements are used.

The global Positioning System (GPS) provides location and time information as long as there are unobstructed lines of sight to four or more GPS satellites. However, when this is not the case the signal may be inaccurate or sometimes even completely blocked. In these situations the Inertial Navigation System (INS) is an appropriate choice for positioning.  An INS has already been proposed in a previous thesis by Erik Andersson and the objective of this thesis is to fuse the GPS with the INS in a proper way. A part of this project is to decide the reliability of the GPS.Three methods for GPS reliability detection have been proposed. One method based on the statistical properties of each of the separate systems, and two methods based on the statistical properties of the residuals between the GPS and INS.

Simulations of simulink models using Kalman filters are often very time-consuming. This problem depends mainly on the fact that the Kalman correction has to be performed at each sample instance through the whole simulation. The goal for this thesis work is to reduce that time-consumption for the filtering part (the integration partis treated in a complementary report) of a simulation. Furthermore a Matlab routine to perform parameter tuning and finally a graphical user interface is developed.The filtering part of the simulation in this thesis is based on an Extended Kalman filter (EKF). The time optimization of this filter considers searching for the possibility to replace the today?s existing Matlab functions that is used to perform the filtering calculations.

At Flight and Systems test Saab AB, a post-processing software is used to process GPS data. A new software by the name GrafNav has been purchased and the purpose of this master thesis therefore became, partly to make a judgment regarding GrafNav?s ability to estimate position, velocity and accuracy, partly to if needed improve the estimate and finally find one or several methods to estimate the position and velocity accuracy. The judgment of GrafNav was performed partly by a comparison to the former post-processing software (PNAV) and partly by a comparison to the airplane?s inertial navigation system (INS).

When a helicopter is flying, the dynamics vary depending on, for example, speed and position. Hence, a time-invariant linear model cannot describe its properties under all flight conditions. It is therefore desirable to update the linear helicopter model continuously during the flight. In this thesis, two different recursive estimation methods are presented, LMS (Least Mean Square) and adaptation with a Kalman filter. The main purpose of the system estimation is to get a model which can be used for feedback control.

This master thesis studies distributed navigation which isa function implemented in a future network based combat information system to improve the accuracy in navigation for combat vehicles in a mechanized battalion, above all in the event of loss of GPS. In the event of loss of the GPS the vehicles obtain dead reckoning performance through the backup system that consists of an odometer and a magnetic compass. Dead reckoning means a drift in the position that makes the accuracy in the navigation worse. The distributed navigation function uses position and navigation data with measurements between the vehicles to estimate the errors and uncertainties in positions, which are used to improve the accuracy in position for the vehicles. To investigate and demonstrate distributed navigation, a simulation environment has been produced in Matlab.

WesDyne TRC AB is a fully owned subsidiary of the global WestinghouseElectric Corporation. WesDyne is a world leader in non-invasive methods for theinspection of nuclear power plants. One of the inspection platforms used to scanfor defects in and around the welds in the reactor vessel is called the T-crawler.The T-crawler operates under water in a radioactive environment by stepping onthe vessel (reactor tank) inner wall using suction cups. The environment is harshand many traditional positioning sensors are not suitable for implementation. Thisthesis investigates the possibility to use a gyroscope and a pressure sensor tocreate an alternative and independent positioning system.

This master thesis is a study on how a rategyro triad, an accelerometer triad, and a magnetometer triad can be integrated into a navigation system, estimating a vehicle?s attitude, i.e. its roll, tipp, and heading angles. When only a rategyro triad is used to estimate a vehicle?s attitude, a drift in the attitude occurs due to sensor errors.

This thesis approaches the problem of estimating the arm angles of an industrial robot with flexibilities in joints and links. Due to cost-cutting efforts in the industrial robots industry, weaker components and more cost-effective structures have been introduced which in turn has led to problems with flexibilities, nonlinearities and friction. In order to handle these challenging dynamic problems and achieve high accuracy this study introduces state observers to estimate the tool position.The observers use measurements of the motor angles and an accelerometer and the different evaluated observers are based on an Extended Kalman filter and a deterministic variant. They have been evaluated in experiments on an industrial robot with two degrees of freedom. The experimental verification shows that the state estimates can be highly accurate for medium frequency motions, ranging from 3-30Hz.

Simulation of different systems can be done using a graphical model description of the system based on block diagrams. The simulation software Matlab/Simulink offers a lot of possibilities to describe a huge number of systems, which can be linear as well as nonlinear. The simulation using Matlab/Simulink is done by integration based on different kinds of numerical methods.When nonlinear Simulink-models are simulated, it is often interesting to apply a Kalman filter to handle the noise that may occur within the process as well as corrupting the measurement. However, application of such a filter leads to an extensive time-consumption despite using fast computers. The reason is that the existing Matlab functions are not constructed for time optimal evaluation.

How to create a more effective filter for the dishwashers of tomorrow..

Den obemannade  helikoptern Skeldar utvecklas av Saab Aeronautics. Saab är intresserade av hur väl sensorfusion kan användas för navigering av en obemannad helikopter. Med sensorfusion kan tillgängliga sensordata fusioneras. Det är mätningar från accelerometer, gyro, GPS och magnetometer som används av olika Kalmanfilter. Olika sensorfusionsmetoder föreslås, däribland Extended Kalman filter (EKF) och ett heuristiskt filter för att visa på hur man intuitivt kan göra för att fusionera mätningar jämfört med det modellbaserade sättet som Kalmanfiltret erbjuder.

This paper describes the work done trying to diminish the noise present in the signal from an IR-detector. The project was aiming towards two different kinds of applications; heat protection and presence detection. The hypothesis was that the signal-noise ratio would decrease if a filter and a proper installation were used for the detector. Some experiments were executed to determine what kind of filter were the most suitable for each application and how these could be installed. After executing the experiments and evaluate the results we recommend to use a specific filter for each application.

IIR-filter är överlägsna traditionellt skapade FIR-filter på många punkter. De klarar sig med en klart lägre ordning för att realisera ett filter med samma egenskaper. En nackdel är däremot att filtret är klart känsligare för koefficient kvantisering och förlorar sina egenskaper när koefficienterna ändras, detta inträffar även om förändringen är liten. Genom den speciella struktur som vågdigitala filter använder sig av kommer man runt detta problem. Det betyder bl.a.

Normally the operational amplifier is used as a component in active filter design. The goal for future components in filter design is large bandwith and the use of low supply voltages. The use of current-mode instead of voltage-mode circuits gets a step closer to the required performance. A component that use current as the signal conveying part is the current conveyor. In this report a comparative study between two current conveyor structures of second generation (CCII) is performed.